Single Cell Analysis of Peripheral TB-Associated Granulomatous Lymphadenitis

We successfully employed a single cell RNA sequencing (scRNA-seq) approach to describe the cells and the communication networks characterizing granulomatous lymph nodes of TB patients. When mapping cells from individual patient samples, clustered based on their transcriptome similarities, we uniformly identify several cell types that known to characterize human and non-human primate granulomas. Whether high or low Mtb burden, we find the T cell cluster to be one of the most abundant. Many cells expressing T cell markers are clearly quantifiable within this CD3 expressing cluster. Other cell clusters that are uniformly detected, but that vary dramatically in abundance amongst the individual patient samples, are the B cell, plasma cell and macrophage/dendrocyte and NK cell clusters. When we combine all our scRNA-seq data from our current 23 patients (in order to add power to cell cluster identification in patient samples with fewer cells), we distinguish T, macrophage, dendrocyte and plasma cell subclusters, each with distinct signaling activities. The sizes of these subclusters also varies dramatically amongst the individual patients. In comparing FNA composition we noted trends in which T cell populations and macrophage/dendrocyte populations were negatively correlated with NK cell populations.

In addition, we also discovered that the scRNA-seq pipeline, designed for quantification of human cell mRNA, also detects Mtb RNA transcripts and associates them with their host cell’s transcriptome, thus identifying individual infected cells. We hypothesize that the number of detected bacterial transcript reads provides a measure of Mtb burden, as does the number of Mtb-infected cells. The number of infected cells also varies dramatically in abundance amongst the patient samples. CellChat analysis identified predominating signaling pathways amongst the cells comprising the various granulomas, including many interactions between stromal or endothelial cells and the other component cells, such as Collagen, FN1 and Laminin,. In addition, other more selective communications pathways, including MIF, MHC-1, MHC-2, APP, CD 22, CD45, and others, are identified as originating or being received by individual immune cell components.

Author Summary

The research conducted describes the cellular composition and communication networks within granulomatous lymph nodes of tuberculosis (TB) patients, employing a single-cell RNA sequencing (scRNA-seq) approach. By analyzing individual patient samples and clustering cells based on their transcriptome similarities, the study reveals several consistent cell types described to be present in both human and non-human primate granulomas. Notably, T cell clusters emerge as abundant in most samples. Additionally, variations in the abundance of B cells, plasma cells, macrophages/dendrocytes, and NK cells among patient samples are observed. Pooling scRNA-seq data from 23 patients enabled the identification of T, macrophage, dendrocyte, and plasma cell subclusters, each displaying distinct signaling activities. Moreover, the study uncovers a surprising capability of the scRNA-seq pipeline to detect Mtb RNA transcripts within host cells, providing insights into individual infected cells and Mtb burden. CellChat analysis unveils predominant signaling pathways within granulomas, highlighting interactions between stromal/endothelial cells and other immune cell components. Moreover, selective communication pathways involving molecules such as Collagen, FN1, Laminin, CD99, MIF, MHC-1, APP and CD45 are identified, shedding light on the intricate interplay within granulomatous lymph nodes during TB infection.